Brain imaging studies have typically focused on localization of function. I will describe recent work on developing a more unified approach to brain function using network models. I will first discuss methodological and conceptual issues involved in characterizing networks from brain imaging data. I will then focus on three key canonical brain networks that reflect strong intrinsically-coupled neural activity (1) an executive control network (ECN) anchored in the dorsolateral prefrontal cortex and posterior parietal cortex, (2) a salience network (SN) anchored in the orbital frontoinsular cortices and the dorsal anterior cingulate (dACC) with robust connectivity to subcortical and limbic structures, and (3) a “default-mode” network (DMN) anchored in the ventromedial prefrontal cortex (VMPFC) and the posterior cingulate cortex (PCC). I will summarize recent research regarding dynamic interactions between these brain networks in the endogenous “resting-state” and during cognitive information processing. When combined with EEG, the temporal dynamics within such brain networks can be investigated with a precision of about 10 milliseconds, yielding new insights into some of the mechanisms underlying cognitive control. The relation between individual variations in intrinsic network connectivity and differences in behavior and cognition will be explored. Finally, I will describe some clinical applications and show how these brain networks are perturbed in Alzheimer’s disease and major depression, and how these investigations inform us in new ways about the neurobiological substrates of each disorder.